All you need to know about our using our transcription service and products.
Welcome to Scribie
Thank you for interest in our products and services. We provide a range of services related to audio/video transcription.
Scribie makes it easy for you to get high-quality transcripts of your audio/video files. Just upload your files, choose automated or manual and pay. Then download the transcripts from your account. The transcripts are provided as Microsoft Word, OpenDocument Text, Adobe PDF and plain text files. The manual transcripts are time-coded (view sample) and prepared by our home-based certified transcribers following a 4-step transcription process which guarantees high accuracy. You can check the transcript as it's being prepared. We also offer free re-review for manual transcripts in case you want any additional changes made.
We have expertise in interview transcription, webinar transcription, podcast transcription, sermon transcription, focus group transcription, lecture transcription and video transcription. We provide general transcription services as well as specialized transcription services for academic transcription, legal transcription, and business transcription.
Our transcripts are high-quality, consistent, well-researched and delivered on time. We research the tough words, acronyms, names of places and other contextual information. Incomprehensible portions are marked with a blank and time-stamped so that you can easily cross-check with the audio file.
We focus on high quality files with American speakers and offer the cheapest rate with very high accuracy for such files. We have been in business since 2008 and transcribed more than 3 millions minutes to date.
Upload or import audio/video files
Choose automated or manual service and pay
Check and download transcript
If you uploaded files as a guest, an account will be created for you with the email address supplied during payment. You can monitor the progress of your files, view the draft transcript, download the transcripts, manage payment methods, order more files from your account
We require payments to be made in advance via credit card and/or PayPal for order amounts of less than $1000. We start working on the file only after the payment has been processed. We also provide free credits for trial runs. Please join our newsletter to avail the credits.
We also support billing/invoicing for larger orders. A contract has to be executed to set up a billing account. Please contact us for additional information.
The turnaround time is indicative and files may be delayed during holidays and weekends or due to excessive load or backlog.
Additional charges may apply for files with non-American accents, poor audio quality, distortions, distant speakers, high background and/or ambient noise. A full refund will be issued if the additional charges is unacceptable, or if the file is un-transcribeable.
Files longer than 3 hours may be delayed. To avoid that please consider splitting the file into two or more parts by trimming it. The split point should be chosen so that the conversation is not interrupted and each part is no longer than 3 hours.
Automated transcripts processed one at a time. We provide an estimate of the time for each file. The delivery make take longer in case our queue is overloaded. Please contact support in case you need to prioritize your file.
The transcript order can be cancelled anytime before 60% completion (as assessed by progress percentage). Any refunds are proportional and are issued after deducting the cost of work already completed on each file. You can choose to send the refund to your payment method which was used or your account credits.
We do not offer refunds after the transcript has been delivered. However, we offer free re-reviews if the transcript quality is unsatisfactory or any additional changes are required. We will redo the transcript as many times as required to achieve the 99% minimum accuracy as promised.
We also do not offer refunds for blank portions in the audio file where there is no spoken audio. We recommend editing the audio file with Audacity or using Trim Audio functionality on the files page.
The Transcript Editor is an online tool which helps you quickly and efficiently check the transcript against the audio file. Click on the audio waveform and the cursor will placed in the corresponding position of the transcript. Conversely, place the cursor in any position of the text and you can play the corresponding audio. Additionally, you can also play from the start of paragraphs, check and edit any in-audibles which are timestamped and marked with a blank, and cycle through blanks. On Google Chrome, audio playback can also be sped up or slowed down to carefully examine difficult audio sections. Various keyboard shortcuts are also available to help you navigate without the mouse.
The Transcript Editor is packed with features and is a huge time saver for assessing the accuracy of the transcript. We strongly recommend that you check the file before downloading and add those last finishing touches before using it for your end purpose.
Please also read The Integrated Editor blog post on how to use this tool and learn about keyboard shortcuts that can help you navigate its functions more efficiently.
Maintaining confidentiality of all transcription files is our highest priority. Access to the files is restricted strictly on a need to know basis. Only our employees and contractors have access to the full audio. All our transcribers, employees and contractors are also bound by a confidentiality clause in the Terms & Conditions. Furthermore, you have full control over your files. Once you delete the files, we remove it permanently from our servers. You can also delete your Scribie.com account anytime.
We also split up the file into smaller parts before we distribute it to our transcribers. That way they only have access to parts of the file. Nobody but our employees and contractors have access to the full audio file. We also use industry standard 256-bit SSL encryption on our website so that all communication between servers and browsers is encrypted.
Please note that we are not currently HIPAA compliant.
Our minimum quality standard is 99% accuracy and the transcripts are delivered only when this criterion is met. Our 4-step transcription process ensures that each file is checked multiple times by different people. More eyes and ears on the file results in a better quality transcript. We also do any research required for technical terms and acronyms. Our Proofreaders are chosen only from the best of certified transcribers. We also have subject matter experts for Legal Transcription, Academic Transcription and Business Transcription as well as accent experts for Indian, African and other non-native accents.
Our philosophy is to keep re-working the file till the minimum quality standard is achieved. We have a dedicated Quality Assurance Team and all the team members are carefully chosen and vetted. They have been trained in the Proofreading and QC process. Our process itself is very structured and with the help of tools can efficiently identify files which need reworking. The proofreading step is repeated with a different proofreader whenever the QC fails and delivered only when the QC passes with minimum 99% accuracy.
Accuracy also depends on the quality of the audio file and the difficulty level. The accuracy may be lower if the file contains background noise, or if the speakers cannot be heard clearly, or if the speakers’ diction or accent is hard to comprehend. Such files may require 1 week to deliver instead of 1 day since it requires extra effort to transcribe accurately. The full amount paid will be refunded if the file is un-transcribeable.
You can also order a free re-review if the transcript quality is not satisfactory after the transcript has been delivered. The re-review order can be placed from your account along with specific instructions you may have. The file will be proofread once more and re-delivered. Re-review orders take 1 business day to process.
In contrast, our competitors follow a 2-step process which has proven to be a failure in the industry. Transcription is a highly skill based task and achieving high accuracy with a 2-step process is very hard to achieve at scale. Our 4-step process is designed to scale and produce high accuracy transcripts consistently.
We do not offer refunds after the transcript has been delivered. However, the transcript order can be canceled at any time until it reaches 60% completion.
The accuracy of automated transcript is highly dependent on the audio quality and the accent of the speakers. The accuracy may be as high as 90% for clean audio files with American speakers. Any audio issue will adversely affect the accuracy. For poor quality files with accented speakers, the accuracy may be as low as 60%.
The automated transcript service is provided 'as-is' without any guarantees. Please use it at your own risk. We do not offer any refunds for automated transcripts.
We offer free re-reviews for manual transcript orders in case you find the transcript quality unsatisfactory after delivery. The re-review order can be placed from your account and along with specific instructions you may have. The file will be proofread once more and re-delivered. Re-review orders take 1 business day to process.
We strive to maintain a high level of quality and take customer feedback very seriously. For each re-review order placed we investigate the root causes and address them to prevent it from reoccurring in the future.
We do not offer refunds after the transcript has been delivered. However, the transcript order can be cancelled at any time until it reaches 60% completion.
The transcript order can be canceled any time before it reaches 60% completion. The refund is processed within 1 business day and the amount depends on the progress. We calculate this by multiplying the amount paid with the progress of the file. Therefore, if the progress is 0% then the full amount will be refunded. Upon cancellation, the draft transcript (for the percentage that has been completed) can be downloaded as a text file from your account.
We do not offer refunds after the transcript has been delivered. However, we offer free re-reviews once the transcript has been delivered.
Subscription plan cancellation Policy
Subscription plans can be cancelled anytime. You will be billed till the end of the current cycle and will continue to enjoy the subscription plan benefits till that time. You will be automatically converted to a Pay as you go customer at the end of the current subscription cycle.
You can share your account with others via our Teams feature. Each team member will be given individual login credentials and will have full access to your account and vice versa. They can upload files, order transcripts, check the transcripts, etc. The payment method added by the first team member is also shared with everyone else and orders can be placed using that. All team members are also cc'ed in all the important email notifications. This feature allows you to share your files securely with another person without sharing your password.
You can add/manage team members from the settings page. There is no upper limit to the number of members.
Account credits are funds which you can add to your Scribie account and use it to pay for transcript orders, automated transcripts or any other payments on Scribie. Your credit card will be charged only once. You can add anywhere from $10 to $15,000 at a time. Account credits can be also be shared with team members and they do not expire. However, account credits cannot be withdrawn or transferred. You can only use it for payments on Scribie.
You can add credits from the settings page. You can also choose to send any refunds to your account credits.
We follow a 4-step manual transcription process. The work is done by our certified home-based freelancers with increasing skill levels at each step. Our philosophy is to keep re-working the transcript till the minimum quality standard of 99% accuracy is achieved. The following is an illustration of our process.
Our online transcription system manages all of the above stages. All our transcribers are home-based freelancers who have gone through our certification process and are paid on an hourly basis for the work done. The system monitors their performance and manages the payments made to them. We provide them with tools to efficiently transcribe files and a comprehensive set of guidelines for transcription. The guidelines are enforced strictly and their performance is graded. The transcripts are prepared segment by segment and quality improves at each stage.
We have re-thought the audio transcription process for the internet-age and designed our audio transcription system from scratch to produce the best quality transcripts at the least cost and effort. Our system and service has a proven track record. We have transcribed more than 3 Millions minutes to date. The following are some unique features which our system provides.
- Accurate measurement of transcript progress. You can track the progress and check the draft transcript anytime.
- Collaborative team to on each file, resulting in higher transcript quality. Even the shortest file is worked upon by more than three people.
- Faster completion time since a number of people can work on different parts of the transcript simultaneously.
- Consistent and repeatable results due to a common set of guidelines.
Our system is unique and can be best described as a Mechanical Turk for audio transcription. A series of posts on our blog describes the Transcription System in detail.
The automated transcripts are prepared by our speech recognition engine. The engine has been built with cutting edge AI algorithms trained our own dataset which we have collected over the years. Because of that, our system tends to perform better on real world data. Our speech recognition engine performs comparably to state-of-the-art systems from Google, Microsoft and Amazon.
Automated transcripts are untouched by humans.
The text of the transcript is broken into paragraphs and each speaker's diction is transcribed on a new paragraph. New paragraphs are started at every change of speaker or at every 3 minutes, whichever is earlier. We also provide audio time-coding and speaker tracking by default. If names of speakers are provided or mentioned in the audio, then they are used. If not, Speaker 1, Speaker 2 and so on are used. For an illustration, please check the sample transcript. For monologues, the speaker name or Speaker 1 is used throughout. Monologues are also broken into paragraphs at every 3 minutes, roughly. Both audio time-coding and speaker tracking can be disabled when ordering or downloading the transcript. The transcript does not include any speech analysis symbols.
The transcripts are clean verbatim by default and all utterances (e.g. 'mm-hmmn', 'uh-huh', 'um', 'uh'), fillers (e.g., 'you know', 'like', 'right', 'so'), interjections and false starts are omitted. Conversely, everything is included when the Strict Verbatim option is enabled.
The spelling style can also be specified along with the order. Currently we support American, British, Australian and Canadian English style spellings. It is automatically chosen based on your current location, if not specified. Please contact us with your requirement for more options.
Custom formatting is provided on request and is charged extra depending on the requirements. The charge depends on the complexity of formatting and the variance from the default format. Please contact us with your requirements for an exact quote.
Transcript File Formats
The transcript files are delivered in the following formats:
- Microsoft Word (.doc)
- Adobe PDF (.pdf)
- OpenDocument Text (.odt)
- Plain text (.txt)
A template file is used for conversion for the first three formats. The plain text file contains only the transcript in UTF-8 encoded text.
Our base rates are applicable only for files with American accent speakers and we charge additional for other accents. We have expertise in British, Australian, Indian, African, various European and Asian accents. We guarantee 99% accuracy for accented speaker files as well. Accented files may require longer time and also incur additional charges.
We can transcribe files recorded in noisy backgrounds such as café's, crowds etc. as long as the main speakers are clear and audible. Files with ambient background noise, e.g. air conditioning hiss, can also be transcribed provided the main speakers can be heard clearly. If the files are untranscribable, a full refund is issued. Such files may also require a longer time to process and also incur the additional charges.
Our rates and turnaround times are applicable only for files with clean audio and American speakers. An additional charge of $0.50/minute to $2.00/minute and a delay may apply for files with the following issues.
|Ambient Noise||hiss, line noise, static, background music/voices|
|Noisy Environment||street, bar, restaurant or other loud noises in background|
|Distant Speakers||faint, distant voices|
|Accented Speakers||British, Australian, Indian, Hispanic, any other non-American|
|Audio Breaks||bad phone line, audio gaps|
|Disturbances||loud typing sounds, rustling, wind howling, breathing sounds|
|Distortion||volume distortion, shrill voices, clipping, artifacts|
|Unclear Speakers||muttering, volume variation, frequent overlaps|
|Echo||reverberation, same voice can be heard twice|
|Quality||low sampling/bit rate, bad conference line, recorded off speakers|
|Diction||slurring, rapid speaking, unnatural pronunciation|
|Muffled||hidden or obstructed microphone, vintage tapes|
|Blank||only music, only background conversation, only non-English|
The charge is NOT automatic and it can be declined. Additional charges are non-negotiable and final. As per our terms of service, our assessment final and binding and any disagreement will lead to immediate cancellation of the order.
As per our data, files with one or more of the above listed issues require around 3 to 4 times the amount of effort to ensure 99% accuracy. We have found that compensating our transcribers for the additional time and effort is the most efficient way to achieve our goal of 99% accuracy since it distributes the effort evenly across the 4 steps of our process. Therefore we have optimized our system for clean files with American speakers, and we provide the best accuracy to price ratio for such files.
Please check the recording tips for recommendations on how to avoid these charges and read our manifesto for the broader reasoning behind this policy.
The audio quality of the recording is the most important factor which affects the transcript accuracy. Good audio files require lesser time and effort to transcribe and therefore our system is designed to favor them. In comparison, poor audio files require around 3 to 4 times more effort. The following simple steps will help you save a lot on the transcription costs and result in highly accurate transcripts.
The following recommendations apply to all types of recordings.
- Smartphones are the best recording devices available today and we recommend using smartphones instead of digital/tape recorders. Smartphones have good microphones and provide built-in apps for recording.
- The first 30 seconds should always contain silence so that it can be used as a noise profile while cleaning the file. We recommend Audacity for cleaning.
- We recommend recording with at least two devices so that there is an alternate version available if required.
- We recommend using a minimum of 128 kbit/s encoding sampling rate with a 16 bit depth.
- Recordings should be done in an environment with least ambient noise. If ambient noise is unavoidable, then we recommend recording 30 seconds of silence at the beginning and using that to clean the recording afterwards.
In-person recordings are recordings made with two or more participants in the same room or place. The following are our recommendations:
- The recorder should be kept equidistant from the participants. For group discussions or focus groups, the optimal arrangement for participants is a circle with the recording device kept at the center.
- The recording device should be kept on a book or some other soft material which can absorb keyboard sounds or any other disturbances. Disturbances sometimes can obscure the speaker diction and should be kept avoided as much as possible.
- For lectures/event recordings where only the main voice is required, edit the audio file afterwards and remove other voices. This will reduce the cost of the transcript as well.
Phone/Online Call Recordings
Phone recordings are interview/conversations over the phone which are recorded off-the-speaker. We recommend using a conferencing service to record phone calls and fall back on recording off-the-speaker. These recommendations also apply for internet phone calls such as Skype, Google Hangouts, WebEx calls, etcetera.
- The recording device should directed towards the speaker of the phone/computer.
- Use high quality speakers, if possible external speakers connected to the phone/computer. Smartphones do not usually have good speakers and should be avoided. We recommend using it only as the recording device as they have good microphones.
- If there is a detectable line noise (buzzing sound on the phone line), try disconnecting and calling again. Line noise is very hard to remove and disorienting for the transcribers.
We write extensively on our blog about how to record good quality files for transcription. The following are some relevant links.
There are other tutorials/videos available online and we recommend that you research about a good recording setup independently as well.
We will include all utterances, e.g. 'mm-hmm', 'uh-huh', 'umm', 'uh' in the transcript. By default transcripts are non-strict verbatim and do not include these words unless necessary. An additional charge of $0.50/minute of audio is applicable for this add-on.
Burnt-In Time Coding
The Burnt-In Time Code in the video frame will be used for time coding the transcript. We require the video files to be uploaded with the time code burnt into the screen. An additional charge of $0.50/minute is applicable for this add-on.
We also provide subtitle files in SubRip (.srt) and WebVTT (.vtt) formats. Only basic versions are supported, without any style markup. The time-code accuracy is +/- 1 seconds. Please contact us if you require subtitle files in other formats. No additional charges apply for this option.
We provide audio time-coding by default. Time-coding refers to the process of adding the running time of the audio before each paragraph. New paragraphs are started at every change of speaker or 3 minutes, whichever is earlier. Each paragraph begins with the audio time which indicates the running time of the audio file at which the particular speaker started speaking for that paragraph. We provide an accuracy of +/- 1 second for the time stamp. For higher accuracies, please contact us.
We also support Burnt-In Time Code (BITC) where the on-screen time code is used. It is an add-on.
We provide speaker tracking by default. Speaker tracking refers to the process of adding the speaker tags before each paragraph. New paragraphs are started at every change of speaker or every 3 minutes, whichever is earlier. Before each paragraph, we add the speaker name, if available, or their order of appearance in the audio, e.g. Speaker 1, Speaker 2 etc. We also provide an option to specify the speaker names before the order is placed. The speaker names can also be changed after the file has been delivered using our Integrated Editor.
We do not have any maximum number of speakers to be tracked. However, it may be difficult to identify the speakers if there are more than four or more speakers in any file. We mark such speakers as S? in the transcript. This issue can be avoided if the speakers speak aloud their names at each turn. Otherwise the speaker tracking is best-effort and we cannot guarantee the correctness.
There is no extra charge for speaker tracking.
We support the following spelling styles.
This option allows you to specify the template file used for document conversion. The following transcript templates available.
|Scribie Single Line Spaced||Contains a title page, Scribie Logo, header and footer, transcript is single line spaced|
|Scribie Double Line Spaced||Same as above, except the transcript is double line spaced|
|Blank Single Line Spaced||Contains only the transcript which is single line spaced|
|Blank Double Line Spaced||Same as above, except the transcript is double line spaced|
The sample transcript has been formatted with Scribie Single Line Spaced template. We also support custom template files. Please contact us with your template file to have it set up.
Our Transcription Progress Tracking feature enables you to track the progress of a transcript from your account at any stage. We provide the completion percentage as well as the draft transcript. The percentage is calculated based on the stage at any point in time. We also provide a draft transcript which shows the changes made. The accuracy of the draft transcript may be lower and may also contain inconsistencies (all of which will be corrected during later stages). The following is the mapping between the percentage and the stage of the process.
Progress Tracking is a unique feature of Scribie. The draft transcript is useful if you need something to work on right away. The Transcript Progress blog post has more details about this feature.
Unique Selling Points
Our transcription service offers features that save you time and money. Transcription is a very labor intensive work and we ensure that you get the best possible transcript.
- Our transcription process ensures that the file is checked multiple times by different people. Even the shortest file is worked on by more than three people. More eyes and ears produce a higher quality transcript.
- You pay a fixed price for the transcript based on the audio length and not the time spent transcribing.
- We provide the best price-to-accuracy ratio for clean files with American speakers.
- We are committed to quality. The additional charges ensure that the transcript is as accurate as possible, regardless of the difficulty level.
- We carefully vet each one of our transcribers and continuously monitor their performance.
- We research uncommon terms, acronyms, etc. and ensure that the transcripts are contextually correct.
- We deliver files on time. If we anticipate a delay then we will inform you well in advance.
- You can monitor the progress and check the Draft Transcript from your account at any stage.
- You can order a free re-review if you are not satisfied with the transcript.
- We are consistent, predictable and accurate. You can plan around us.
- As a company, we take confidentiality very seriously and we handle your files with utmost care.
- We respond to your questions in a timely manner and continue supporting you even after the file is delivered.
The following is a sample audio and the corresponding transcript. More sample transcripts are available on our blog.
We provide phone, live chat and email support on all weekdays. Our customer support team is well trained and responsive. We ensure that all questions and issues are addressed promptly. Please do not hesitate to reach out to us.